How do you remove the sound of breath?

How do you remove the sound of breath? I tried using the noise reduction effect, it removed the background noise and now the breath is much clearer.

Thank you for any assistance.

If it’s not very bad, the best way is to do it manually, with volume automation. Just pull the volume down when there are breath noises. This keeps the sound natural and doesn’t create any artefacts. It also lowers noise.

See the envelope tool to discover how to do it:

http://manual.audacityteam.org/man/envelope_tool.html

As a sound editor, you really need to know this tool anyway. It’s probably the most used tool in audio editing, before eq, compression or whatever.

If it’s bad, consider posting a sample here. For a how-to on uploading samples to the forum, have a look here:

https://forum.audacityteam.org/t/how-to-attach-files-to-forum-posts/24026/1

After reading the post by Cyrano, I began experimenting with the envelope tool to remove breathing. MIRACLE, this worked flawlessly for me and will save me hours of editing. After removing the breaths, I exported the track as a 32 it WAV file, when I imported the track the envelope was gone, just what I was hoping would happen.

Need to find out if ACX accepts WAV for a final project.

Thanks Ctrano

If you are going to adjust each breath individually (that is to say you don’t apply some sort of noise reduction effect to the entire track at once), then I can save you even more time than using the envelope tool. In fact, I believe the result is better.

My process is to:

  1. Select the breath with your cursor
  2. Reduce amplification to a New Peak Amplitude between -30dB and -40dB
  3. Find the next breath and press CTRL-R (the repeat function using your keyboard)
  4. Continue this process for every breath

Note: The CTRL-R (or Command-R on a Mac) will repeat the previous effect. So if you apply a different effect then you’ll have to go back to Step 2.

The reason I reduce the amplitude to between -30dB and -40dB is because it keeps the pace (better than just deleting it) and reduces the obnoxiousness of a huge in-breath.

Try it. I think you’ll be sold on the process.

Steve Stewart, Podcast Editor

Need to find out if ACX accepts WAV for a final project.

They don’t. They insist on MP3 submission.

http://www.acx.com/help/acx-audio-submission-requirements/201456300

Scroll down in this post to “There’s a trick to submission.”

Koz

Please note that even if you do pass the automated ACX Robot at the beginning of the submission (similar to ACX-Test here on the forum) Human Quality Control is still going to bounce it if they can find pulsing or odd background noises or other evidence of excessive processing. They would call it distracting.

They would rather you didn’t do anything to your production past making the volume come out right. This is a straight copy from their instruction web page.

You are manually gating to get rid of breathing sounds.

We found some restrained corrections could be slipped past if you didn’t go overboard. Maybe some breath sounds would be OK if you’re not asthmatic or chronically short of breath. One of my favorite readers has a very odd reading style and doesn’t make any effort to cover it up.

Koz

Please excuse this if it sounds naive, but I’m just learning Audacity.
@SteveStewartMe, you say to “Reduce amplification to a New Peak Amplitude between -30dB and -40dB.” I go into Effect, Amplify, and then do I plug in a number between -30 and -40 in “Amplification” or “New Peak Amplitude?”

Thanks

I’d love to find something which can be automated across an entire track. That said, I’ve had a lot of success using the Amplify effect and reducing the amplitude by 6dB. The envelope tool is fantastic for doing a lot of experimentation in a short period of time, but for breathing and lip smacking, which at least for me is usually at a consistent amplitude, Amplify works excellently. I have a preset saved and then use Ctrl+R after the first use. Pretty efficient and easy to use.

I understand all of the reticence expressed here about using noise gates. But I have had some success using Reagate:
http://www.reaper.fm/reaplugs
Some people seem to have had difficulties in installing these VST plugins, as did I at first. So here’s how I got it to work (I should say that I wanted only Reagate but this should work for any of the others downloaded in the package): First download and install the 32 bit version of reaper/reaplugs found at the above address. this will be installed in VST Plugins; open the VST file and find the ‘standalone reagate dll’, next click and drag this into Audacity Plugins. when you open a file in Audacity, you should now be able to access this VST plugin; Effects, VST, reagate.
There is a very useful video showing how to use and tweak things in Reagate @:
https://www.youtube.com/watch?v=fYmb_LMSrd4
It’s not a perfect solution, but with some tweaking it’s pretty good.
'Hope this is of some use.

The Nyquist “Noise Gate” plug-in works perfectly well, and is a native format for Audacity. The reticence expressed was about using any Noise Gate, as ACX specifically say not to use one.

There is another process to cover up breath and mouth noises (my particular talent). You can use Punch In and cover up the bad sound with natural background sound or room tone. This can help if you have pumping or breathing background sounds when you cut or over-reduce the error.

https://manual.audacityteam.org/man/tutorial_punch_in_repair_of_recordings.html

A note about ACX mastering. Their goal is you telling a fascinating story over cups of tea. They want the listener to be able to ignore the process in the middle and concentrate on the story (which they paid for). That’s hard to do when the voice or spaces in the middle change over the course of the presentation.

There is a way to use Noise Reduction. If you have a Close But No Cigar presentation because of noise, you can apply “Noise Reduction of the Beast” (6, 6, 6) to the chapter. It will get significantly quieter and nobody can tell what you did. Most demands not to use Noise Reduction are aimed at performers trying to turn complete trash into a chapter submission. That’s not going to work and will cause wine glass voice distortions.

Noise Reduction doesn’t work on sounds that change, such as breathing and tongue ticks. Its neighborhood is microphone hiss and fan or air conditioning wind noises.

Your chapters have to match. So whatever you do to one should be done to them all.

Koz